Abstract Structures show their health information through various forms during their service. Acoustic emission (AE) sensors can capture the relevant information, but fail to realize their value fully. AE, data mining (DM) and cloud computing have promoted the development of centralized management of structures health management (SHM) but accompanied by the slow response, high requirement for computing capacity. This paper puts forward an innovative concept combining AE, DM on fog computing and centralized management to improve these problems. Firstly, raw data are collected by the AE sensors system and transformed into fog node for DM, including removing useless data, consolidating, organizing, and visualizing the result. Then the concise results will be integrated into a data center for centralized management and visualization. This idea is verified by a case study. It includes both a systematic experiment and a real-world application. The test conducts two groups of beams according to actual bridges and gives two parameters analysis of the experiment data to find the characteristics between signals and performance. Then the results are applied into two real bridges for practical application and detect a critical point. It not only explains how this new concept works but also has practical significance for engineering managers.